Perforin is activated by a proteolytic cleavage during biosynthesis which reveals a phospholipid-binding C2 domain

Chronic active Epstein-Barr virus infection associated with mutations in perforin that impair its maturation

Chronic active Epstein-Barr virus infection (CAEBV) is a rare disease in which previously healthy persons develop severe, life-threatening illness. Mutations in the perforin gene have been found in familial hemophagocytic lymphohistiocytosis, which shares some features with CAEBV. We studied a patient who died at age 18, 10 years after the onset of CAEBV. The patient had high titers of antibodies to EBV, EBV RNA in lymph nodes, T-cell lymphoproliferative disease, and hemophagocytic lymphohistiocytosis. DNA sequencing showed novel mutations in both alleles of the perforin gene that resulted in amino acid changes in the protein. The quantity of the native form of perforin from the patient's stimulated peripheral blood mononuclear cells (PBMCs) was extremely low and immunoblotting showed accumulation of an uncleaved precursor form of perforin. Stimulated PBMCs from the patient were defective for Fas-independent cytotoxicity. These data imply that mutations in this patient resulted in reduced perforin-mediated cytotoxicity by his lymphocytes. This is the first case in which perforin mutations have been shown to result in accumulation of the uncleaved, immature form of perforin. Mutations in the perforin gene are associated with some cases of CAEBV with hemophagocytic lymphohistiocytosis.

Adaptor protein 3–dependent microtubule-mediated movement of lytic granules to the immunological synapse

Hermansky-Pudlak syndrome (HPS) is a rare autosomal recessive disease characterized by platelet defects and oculocutaneous albinism. Individuals with HPS type 2 (HPS2) lack the cytosolic adaptor protein 3 (AP-3) involved in lysosomal sorting, and are also immunodeficient. Here we characterize an HPS2 mutation and demonstrate that AP-3 deficiency leads to a loss of cytotoxic T lymphocyte (CTL)-mediated cytotoxicity. Although the lysosomal protein CD63 was mislocalized to the plasma membrane, perforin and granzymes were correctly localized to the lytic granules in AP-3-deficient CTLs. However, the lytic granules of AP-3-deficient CTLs were enlarged and were unable to move along microtubules and dock within the secretory domain of the immunological synapse. These data show that AP-3 is essential for polarized secretion from CTLs.

The ABCs of granule-mediated cytotoxicity: new weapons in the arsenal

Granule exocytosis is the main pathway for the immune elimination of virus-infected cells and tumour cells by cytotoxic T lymphocytes and natural killer cells. After target-cell recognition, release of the cytotoxic granule contents into the immunological synapse formed between the killer cell and its target induces apoptosis. The granules contain two membrane-perturbing proteins, perforin and granulysin, and a family of serine proteases known as granzymes, complexed with the proteoglycan serglycin. In this review, I discuss recent insights into the mechanisms of granule-mediated cytotoxicity, focusing on how granzymes A, B and C and granulysin activate cell death through caspase-independent pathways.

Delivering the kiss of death

Cytotoxic T lymphocytes and natural killer cells kill their targets by secreting specialized granules that contain potent cytotoxic molecules. Through the study of rare immunodeficiency diseases in which this granule pathway of killing is impaired, proteins such as Rab27a have been identified as components of the secretory machinery of these killer cells. Recent evidence suggests that the destruction of activated lymphocytes through granule-mediated killing may be an important mechanism of immunological homeostasis. Although the process by which this occurs is not yet known, it is possible that events taking place at the immunological synapse may render the killer cell susceptible to fratricidal attack by other killer cells.

Single-cell perforin and granzyme expression reveals the anatomical localization of effector CD8+ T cells in influenza virus-infected mice

Influenza virus infection activates cytolytic T lymphocytes (CTL) that contribute to viral clearance by releasing perforin and granzymes from cytoplasmic granules. Virus-specific, perforin-dependent CD8(+) CTL were detected in freshly isolated cells from the mouse lung parenchyma but not from the mediastinal lymph nodes (MLN), where they are primed, or from the spleen during primary influenza virus infection. To determine whether this difference was due to the low frequency or incomplete maturation of effector CTL in MLN, we measured expression of perforin, granzymes A, B, and C, and IFN-gamma mRNAs in CD8(+) populations and single cells immediately after isolation from virus-infected mice. Quantitative PCR revealed significant expression of perforin, granzyme A, granzyme B, and IFN-gamma in activated CD8(+) cells from MLN, spleen, and lung parenchyma. Granzyme C expression was not detected. Individual activated or nucleoprotein peptide/class I tetramer-binding CD8(+) cells from the three tissues expressed diverse combinations of perforin, granzyme, and IFN-gamma mRNAs. Although cells from lung expressed granzymes A and B at higher frequency, each of the tissues contained cells that coexpressed perforin with granzymes A and/or B. The main difference between MLN and lung was the elevated frequency of activated CD8(+) T cells in the lung, rather than their perforin/granzyme expression profile. The data suggest that some CTL mature into perforin/granzyme-expressing effector cells in MLN but reach detectable frequencies only when they accumulate in the infected lung.

Natural killer-like cells in the sheep: functional characterization and regulation by pregnancy-associated proteins

Natural killer (NK) cells represent an important component of the innate immune system. In ruminants there are few reports regarding presence or characterization of NK cells. Although absence of expression of major histocompatibility complex proteins on ovine trophoblast makes it potentially a target for NK cells, little is known about regulation of NK cells by products of pregnancy in sheep. Objectives of the present study were to determine whether cells with characteristics of NK cells exist in preparations of ovine peripheral blood lymphocytes (PBL) and endometrial epithelial cells (EEC) and to determine regulation of such cells by two pregnancy-associated molecules with immunoregulatory properties (ovine uterine serpin [OvUS] and interferon-tau [IFN-tau]). Ovine PBL and EEC lysed a putative NK target cell, the BHV-1 infected D17 cell, and lysis by both types of cells was neutralized by antibody against a molecule called function-associated molecule (FAM) expressed on NK cells of several species. Moreover, inhibitors that interfere with perforin-mediated lysis blocked NK-like activity of PBL. The NK-like lytic activity of PBL and EEC was inhibited by OvUS, whereas ovine and bovine IFN-tau significantly enhanced NK-like activity of PBL. In conclusion, NK-like activity present in preparations of ovine PBL and EEC is mediated by FAM(+) cells, is dependent on processes that involve perforin processing, and is regulated by OvUS and IFN-tau. Inhibition of NK-like activity of PBL and EEC by OvUS is consistent with a role for OvUS in protecting the conceptus from maternal cytotoxic lymphocytes. Stimulation of lysis by IFN-tau implies the existence of other inhibitory mechanisms during early pregnancy to prevent NK cell-mediated destruction of the conceptus.

Xenogeneic human NK cytotoxicity against porcine endothelial cells is perforin/granzyme B dependent and not inhibited by Bcl-2 overexpression

Because of organ shortages in clinical allotransplantation, the potential of pig-to-human xenotransplantation is currently being explored showing a possible critical role for natural killer (NK) cells in the immune response against xenografts. Therefore, we analyzed the cytotoxic pathways utilized by human natural killer cells (hNK) against porcine endothelial cells (pEC). Transmission electron microscopy of pEC cocultured with hNK cells showed both apoptotic and necrotic cell death, whereas soluble factors such as Fas ligand or TNFalpha did not induce apoptosis in pEC. NK lysis of pEC was abrogated by concanamycin A and ammonium chloride, reagents inhibiting the perforin/granzyme B (grB) pathway, but only partially blocked by caspase inhibition with z-VAD-fmk. Overexpression of bcl-2 protected pEC against apoptosis induced by staurosporine or actinomycin D, but failed to prevent hNK cell-mediated lysis. In conclusion, pEC are lysed in vitro by hNK cells via the perforin/grB pathway and are not protected from NK lysis by overexpression of bcl-2.

Lymphocyte-mediated cytotoxicity

Virtually all of the measurable cell-mediated cytotoxicity delivered by cytotoxic T lymphocytes and natural killer cells comes from either the granule exocytosis pathway or the Fas pathway. The granule exocytosis pathway utilizes perforin to traffic the granzymes to appropriate locations in target cells, where they cleave critical substrates that initiate DNA fragmentation and apoptosis; granzymes A and B induce death via alternate, nonoverlapping pathways. The Fas/FasL system is responsible for activation-induced cell death but also plays an important role in lymphocyte-mediated killing under certain circumstances. The interplay between these two cytotoxic systems provides opportunities for therapeutic interventions to control autoimmune diseases and graft vs. host disease, but oversuppression of these pathways may also lead to increased viral susceptibility and/or decreased tumor cell killing.

The genes for perforin, granzymes A-C and IFN-gamma are differentially expressed in single CD8(+) T cells during primary activation

Here we show that the genes for perforin, the three major T cell granzymes (A-C) and IFN-gamma are differentially expressed during primary activation of naive CD8(+) T cells, kinetically and at the single-cell level. When CD44(low)CD62L(high)CD8(+) lymph node T cells were activated with IL-2 and immobilized antibodies to CD3, CD8 and CD11a, expression of perforin, granzyme B and IFN-gamma mRNAs was induced by day 2, and increased in parallel with perforin-dependent cytolytic activity. Granzyme C and A transcripts were not detected until 1 and 3 days later respectively. Single-cell PCR showed that expression frequencies rose in parallel with total levels of each mRNA, but that individual cells expressed diverse combinations of perforin, granzyme A-C and IFN-gamma mRNAs. These expression patterns indicated that the delayed expression of granzymes A and C was not due to late activation of distinct cell subpopulations. Statistical analysis of the data suggested that each gene was differentially regulated at the single-cell level. Individual naive CD8(+) T cells gave rise over 7 days to clones that expressed all five products at the clonal level, but also expressed diverse combinations at the single-cell level. We conclude that, during primary activation, CD8(+) T cells progressively acquired the ability to express most or all of these genes, and that the variable expression patterns observed among single cells within clones and populations reflected transient rather than heritable differences in expression profile.

Functional consequences of perforin gene mutations in 22 patients with familial haemophagocytic lymphohistiocytosis

Familial haemophagocytic lymphohistiocytosis (FHL), an inherited form of haemophagocytic lymphohistiocytosis (HLH) syndrome, is characterized by the overwhelming activation of T lymphocytes and macrophages invariably leading to death in the absence of treatment. FHL is a heterogeneous autosomal recessive disorder, with one known causative gene which codes for perforin, a cytotoxic effector protein. In this study, we have characterized the genotype and phenotype of 14 unrelated families with perforin deficiency. Four new missense mutations of the perforin gene were identified. In every case, perforin gene mutations led to undetectable intracellular perforin expression within cytotoxic cells, while some residual T-cell cytotoxic activity could be associated with certain missense mutations. Clinical and biological analyses did not differentiate between patients with nonsense or missense mutations, although age at diagnosis, which tended to be similar within members of the same family, was delayed in patients from two families belonging to the second group. In one case, consequences of perforin deficiency, diagnosed at birth, could be assessed prior to onset of clinical manifestations. No evidence for T-cell activation could be shown, suggesting that an exogenous event is required to trigger the disease manifestation. Control assessment of perforin expression and cytotoxic assays by lymphocytes from young children led to the conclusion that perforin content of natural killer cells could be a reliable diagnostic test at any age. Altogether, these data enabled a better characterization of perforin deficiency and its consequences, and defined reliable diagnostic tools.

Cytotoxic cell granule-mediated apoptosis: perforin delivers granzyme B-serglycin complexes into target cells without plasma membrane pore formation

The mechanism underlying perforin (PFN)-dependent delivery of apoptotic granzymes during cytotoxic cell granule-mediated death remains speculative. Granzyme B (GrB) and perforin were found to coexist as multimeric complexes with the proteoglycan serglycin (SG) in cytotoxic granules, and cytotoxic cells were observed to secrete exclusively macromolecular GrB-SG. Contrary to the view that PFN acts as a gateway for granzymes through the plasma membrane, monomeric PFN and, strikingly, PFN-SG complexes were shown to mediate cytosolic delivery of macromolecular GrB-SG without producing detectable plasma membrane pores. These results indicate that granule-mediated apoptosis represents a phenomenon whereby the target cell perceives granule contents as a multimeric complex consisting of SG, PFN, and granzymes, which are, respectively, the scaffold, translocator, and targeting/informational components of this modular delivery system.

Role of electrostatic and hydrophobic interactions in Ca(2+)-dependent phospholipid binding by the C(2)A-domain from synaptotagmin I

Most C(2)-domains bind to phospholipid bilayers as a function of Ca(2+). Although phospholipid binding is central for the normal functions of C(2)-domain proteins, the precise mechanism of phospholipid binding is unclear. One of the key questions is whether phospholipid binding by C(2)-domains is primarily governed by electrostatic or hydrophobic interactions. We have now examined this question for the C(2)A-domain of synaptotagmin I, a membrane protein of secretory vesicles with an essential function in Ca(2+)-triggered exocytosis. Our results confirm previous data showing that Ca(2+)-dependent phospholipid binding by the synaptotagmin C(2)A-domain is exquisitely sensitive to ionic strength, suggesting an essential role for electrostatic interactions. However, we find that hydrophobic interactions mediated by exposed residues in the Ca(2+)-binding loops of the C(2)A-domain, in particular methionine 173, are also essential for tight phospholipid binding. Furthermore, we demonstrate that the apparent Ca(2+) affinity of the C(2)A-domain is determined not only by electrostatic interactions as shown previously, but also by hydrophobic interactions. Together these data indicate that phospholipid binding by the C(2)A-domain, although triggered by an electrostatic Ca(2+)-dependent switch, is stabilized by a hydrophobic mechanism. As a result, Ca(2+)-dependent phospholipid binding proceeds by a multimodal mechanism that mirrors the amphipathic nature of the phospholipid bilayer. The complex phospholipid binding mode of synaptotagmins may be important for its role in regulated exocytosis of secretory granules and synaptic vesicles.

Electrostatic control of the membrane targeting of C2 domains

Many proteins involved in signal transduction and vesicle trafficking contain C2 domains whose membrane association is often regulated by calcium. Here, finite-difference Poisson-Boltzmann calculations are used to describe the electrostatic interactions between C2 domains of known structure and phospholipid membranes. The results explain how calcium binding can drive the association of some C2 domains to negatively charged membranes and others to neutral, zwitterionic membranes. Nonspecific electrostatic interactions are shown to be a general feature of many C2 domains of known structure, including the calcium-independent C2 domain of the PTEN tumor suppressor.

Three-dimensional structure of the synaptotagmin 1 C2B-domain: synaptotagmin 1 as a phospholipid binding machine

Synaptotagmin 1 probably functions as a Ca2+ sensor in neurotransmitter release via its two C2-domains, but no common Ca2+-dependent activity that could underlie a cooperative action between them has been described. The NMR structure of the C2B-domain now reveals a beta sandwich that exhibits striking similarities and differences with the C2A-domain. Whereas the bottom face of the C2B-domain has two additional alpha helices that may be involved in specialized Ca2+-independent functions, the top face binds two Ca2+ ions and is remarkably similar to the C2A-domain. Consistent with these results, but in contrast to previous studies, we find that the C2B-domain binds phospholipids in a Ca2+-dependent manner similarly to the C2A-domain. These results suggest a novel view of synaptotagmin function whereby the two C2-domains cooperate in a common activity, Ca2+-dependent phospholipid binding, to trigger neurotransmitter release.

Full-length cloning and 3'-terminal portion expression of human perforin cDNA

BACKGROUND

Perforin (also known as pore-forming protein, PFP) is one of the main effector molecules which natural killer cells (NK) and cytotoxic T lymphocytes (CTL) utilize to kill their targets both in vivo and in vitro. We report the full length of human perforin cDNA, which was cloned from liver tissue.

RESULTS

Sequencing analysis showed that there were discrepancies of four nucleotides and three amino acids compared with previously published sequence of human PFP. The cDNA fragment was then inserted into fusion protein expressive vector pGEX-2T to construct a recombinant expressive plasmid. The C-terminal truncated 125 amino acids polypeptide (410-534aa) of human perforin (hPFP-C) was selectively expressed in a form of fusion protein. Under the induction of IPTG, GST/hPFP-C fusion protein was expressed in E. coli BL21 (DE3). The fusion protein GST/hPFP-C was purified by affinity chromatography with glutathione agarose. The recombinant hPFP-C obtained by thrombin cleavage showed a significant hemolytic activity when tested with rabbit erythrocytes.

CONCLUSION

These results suggest that the domain responsible for lytic function lies not only in the N-terminal portion but also in the C-terminal portion of perforin molecule. The recombinant hPFP-C protein will be useful as a highly purified biological factor for immunological, pathological and structural studies.

Granule-dependent killing of Toxoplasma gondii by CD8+ T cells

Immunization of mice with live bradyzoites of a low-virulent Beverley strain of Toxoplasma gondii has been shown to increase CD8+ T-cell mediated immunity against a highly virulent RH strain. We found that preimmunization with an RH homogenate further enhanced this immunity. Using this model, we investigated the mechanism of CD8+ T-cell mediated protection against T. gondii infection. Splenic cells from mice immunized with RH homogenate and live bradyzoites stimulated apoptosis of RH-infected J774A.1 macrophages in vitro, and at the same time, the immunization significantly suppressed the proliferation of parasites within macrophages, as assessed by measuring 3H-uracil uptake by the parasites. Splenic cells from the immunized mice produced larger amounts of interferon-gamma (IFN-gamma) than did naive splenic cells; however, the production of nitric oxide (NO) by RH-infected macrophages was not enhanced. The elimination of CD8+ T cells from splenic cells significantly reduced their inhibitory action on parasite proliferation as well as their cytotoxic activity against RH-infected macrophages, but it did not affect the production of IFN-gamma. Treatment of CD8+ T-enriched splenic cells from the immunized mice with concanamycin A, but not an anti-Fas ligand monoclonal antibody, significantly reduced their anti-proliferative and killing capabilities, suggesting that the CD8+ T cells induced by immunization with RH antigen and live bradyzoites of the Beverley strain may exert protection against T. gondii infection at least in part through granule-dependent cytotoxic activities.

Inhibition of human NK cell-mediated cytotoxicity by exposure to ammonium chloride

Ammonium-chloride-containing solutions (AC) are routinely used to lyse red blood cells during preparation of PBMC. Although exposure to AC has been described to affect the ultrastructural appearance of large granular lymphocytes and to temporarily inhibit cytolytic activity of PBMC preparations, the cellular basis of this phenomenon has not been studied. Here, the inhibitory effect of AC on human CTL and NK-mediated cytotoxicity has been analyzed in 4-h 51Cr-release assays. The results show that NK killing of K562 leukemia cells and xenogeneic endothelial cells is inhibited by AC exposure. The effect is dose-dependent and reversible, because recovery of cytotoxicity is observed within 15 h of re-culturing. AC does not reduce the viability of NK cells and the inhibitory effect is not mediated by the exhaustive release of granzymes upon AC treatment. In contrast, antigen-specific CTL killing of EBV-transformed B-lymphoblastoid cell lines and xenogeneic PHA lymphoblasts was less sensitive to AC and data are presented suggesting that FasL-induced apoptosis is not inhibited by AC. In conclusion, perforin-mediated NK killing is AC-sensitive whereas CTL killing and FasL-mediated killing appear to be AC-resistant. Therefore, AC represents a powerful tool to study different mechanisms of cell-mediated cytotoxicity and may be helpful in assessing antigen-specific CTL cytotoxicity without the influence of NK cell-mediated background killing.

An unusual C(2)-domain in the active-zone protein piccolo: implications for Ca(2+) regulation of neurotransmitter release

Ca(2+) regulation of neurotransmitter release is thought to require multiple Ca(2+) sensors with distinct affinities. However, no low-affinity Ca(2+) sensor has been identified at the synapse. We now show that piccolo/aczonin, a recently described active-zone protein with C-terminal C(2)A- and C(2)B-domains, constitutes a presynaptic low-affinity Ca(2+) sensor. Ca(2+) binds to piccolo by virtue of its C(2)A-domain via an unusual mechanism that involves a large conformational change. The distinct Ca(2+)-binding properties of the piccolo C(2)A- domain are mediated by an evolutionarily conserved sequence at the bottom of the C(2)A-domain, which may fold back towards the Ca(2+)-binding sites on the top. Point mutations in this bottom sequence inactivate it, transforming low-affinity Ca(2+) binding (100-200 microM in the presence of phospholipids) into high-affinity Ca(2+) binding (12-14 microM). The unusual Ca(2+)-binding mode of the piccolo C(2)A-domain reveals that C(2)-domains are mechanistically more versatile than previously envisaged. The low Ca(2+) affinity of the piccolo C(2)A-domain suggests that piccolo could function in short-term synaptic plasticity when Ca(2+) concentrations accumulate during repetitive stimulation.

Synaptotagmin I functions as a calcium regulator of release probability

In all synapses, Ca2+ triggers neurotransmitter release to initiate signal transmission. Ca2+ presumably acts by activating synaptic Ca2+ sensors, but the nature of these sensors--which are the gatekeepers to neurotransmission--remains unclear. One of the candidate Ca2+ sensors in release is the synaptic Ca2+-binding protein synaptotagmin I. Here we have studied a point mutation in synaptotagmin I that causes a twofold decrease in overall Ca2+ affinity without inducing structural or conformational changes. When introduced by homologous recombination into the endogenous synaptotagmin I gene in mice, this point mutation decreases the Ca2+ sensitivity of neurotransmitter release twofold, but does not alter spontaneous release or the size of the readily releasable pool of neurotransmitters. Therefore, Ca2+ binding to synaptotagmin I participates in triggering neurotransmitter release at the synapse.

Enhancement of human cord blood CD34+ cell-derived NK cell cytotoxicity by dendritic cells

NK cells and dendritic cells (DCs) are both important in the innate host defense. However, the role of DCs in NK cell-mediated cytotoxicity is unclear. In this study, we designed two culture systems in which human cord blood CD34(+) cells from the same donor were induced to generate NK cells and DCs, respectively. Coculture of the NK cells with DCs resulted in significant enhancement of NK cell cytotoxicity and IFN-gamma production. However, NK cell cytotoxicity and IFN-gamma production were not increased when NK cells and DCs were grown together separated by a transwell membrane. Functional studies demonstrated that 1) concanamycin A, a selective inhibitor of perforin/granzyme B-based cytolysis, blocked DC-stimulated NK cytotoxicity against K562 cells; and 2) neutralizing mAb against Fas ligand (FasL) significantly reduced DC-stimulated NK cytotoxicity against Fas-positive Jurkat cells. In addition, a marked increase of FasL mRNA and FasL protein expression was observed in DC-stimulated NK cells. The addition of neutralizing mAb against IL-18 and IL-12 significantly suppressed DC-stimulated NK cell cytotoxicity. Neutralizing IFN-gamma Ab almost completely inhibited NK cell cytotoxicity against Jurkat cells. These observations suggest that DCs enhance NK cell cytotoxicity by up-regulating both perforin/granzyme B- and FasL/Fas-based pathways. Direct interaction between DCs and NK cells is necessary for DC-mediated enhancement of NK cell cytotoxicity. Furthermore, DC-derived IL-18 and IL-12 were involved in the up-regulation of NK cell cytotoxicity, and endogenous IFN-gamma production plays an important role in Fas-mediated cytotoxicity.

Functional defects of NK cells treated with chloroquine mimic the lytic defects observed in perforin-deficient mice

NK cells are the primary effectors mediating acute rejection of incompatible bone marrow cell grafts. To reduce rejection, we evaluated the ability of chloroquine (CHQ) to prevent perforin-dependent NK cell activity. Perforin is a key cytotoxic component released from the lytic granules of activated NK cells. Generation of functional perforin requires an acidic protease activity that occurs in the secretory, lytic lysosomes. Our hypothesis was that CHQ, a lysosomotropic reagent, would raise the pH of the acidic compartment in which perforin is processed and thereby block perforin maturation and cytotoxicity. We have measured NK cytotoxicity in vivo by clearance of YAC-1 tumor cells from the lungs and by rejection of incompatible bone marrow transplants and in vitro by cytolysis of YAC-1 and Jurkat cells. The engraftment of bone marrow cells was monitored by recolonization of the spleen with hemopoietic cells from transplants of MHC class I-deficient bone marrow cells into lethally irradiated recipient mice. Transplant rejection was compared in two inbred strains of mice: 129, which apparently use perforin-dependent cytotoxicity, and C57BL/6, in which rejection can be perforin-independent. CHQ treatment reduced NK cell activity in 129 mice in which perforin is important for mediating rejection. CHQ affected the fraction of NK cell cytolysis that was Fas independent. In addition, we found that CHQ prevents perforin processing by LAK cells in vitro. These data indicate that CHQ may impair rejection of incompatible bone marrow transplants and other functions mediated by NK and cytotoxic T cells.

Anti-Tumor CC49-zeta CD4 T cells possess both cytolytic and helper functions

The authors report that the nature of the T-cell-receptor--derived signal in normal CD4+ T cells can induce interleukin-2 (IL-2) secretion or perforin-mediated cytolytic activity. Normal human T cells were genetically modified to express the tumor antigen specific chimeric immune receptor, CC49-zeta. The CC49-zeta chimeric immune receptor is comprised of the intracellular signaling domains of the TCR CD3zeta protein fused to the single chain scFv of the humanized CC49 antibody, which binds the pan-adenocarcinoma tumor antigen TAG-72. Patient-specific T cells genetically modified to express the CC49-zeta receptor have been used in patients with colon cancer. The authors report that both CD4 and CD8 T cells expressing the CC49-zeta receptor mediated the major histocompatibility complex-unrestricted lysis of TAG-72--expressing tumor cells with comparable efficiency. However, although the CC49-zeta receptor mediated target cell lysis, it did not support the production of IL-2, even in the presence of CD28 stimulation. Robust IL-2 secretion and T-cell proliferation were observed when the same CD4 CC49-zeta T cells were stimulated through the CD28 receptor and endogenous T-cell receptor. These results indicate that CD4 T lymphocytes possess the capacity to act as both cytolytic and helper T cells and that this difference in effector function is controlled by the nature of the T-Cell receptor--derived signals.

A role for P-glycoprotein in regulating cell death

P-glycoprotein (P-gp) is an energy dependent drug pump responsible for multidrug resistance (MDR) in human cancers. While it is irrefutable that P-gp can efflux xenobiotics out of cells, the biological function of P-gp in multicellular organisms has yet to be firmly established. The question of what, if anything, P-gp does when not effluxing drugs has been raised by recent reports indicating that P-gp may regulate apoptosis, chloride channel activity, cholesterol metabolism and immune cell function. There is now a lively debate regarding the possible role of P-gp in regulating cell differentiation, proliferation and survival.

Nitric oxide synthase-2 and expression of perforin in uterine NK cells

In human, mouse, and rat pregnancy, maternal NK cells accumulate and differentiate at implantation sites. These cells, termed uterine NK (uNK) cells, express NO synthase (NOS)-2 and develop cytolytic molecules such as perforin and granzymes during differentiation in situ. In this study, relationships between expression of the NOS-2 gene, uNK cell population density and tissue distribution, and synthesis of perforin were investigated. Uteri from wild-type (WT) and NOS-2-/- mice were collected at gestation days (g.d.) 8, 10, 12, 14, and 16 (n, >2/g.d.). Histochemical staining failed to reveal any differences between the population densities or tissue distributions of uNK cells in WT and NOS-2-/- uteri at any stage of gestation. By contrast, immunohistochemical staining with anti-perforin Abs demonstrated significantly fewer perforin-positive uNK cells in two uterine compartments of NOS-2-/- mice in comparison to the same compartments in WT mouse uteri. Perforin-positive uNK cells were reduced in NOS-2-/- metrial glands at g.d. 8, 10, and 12 and in decidua basalis at g.d. 12 (p < 0.05). Analysis of perforin protein by immunoblotting confirmed this observation. Northern blot hybridization studies showed that loss of perforin protein in NOS-2-/- mice was accompanied by decreased steady-state levels of perforin mRNA. These results demonstrate that migration of uNK cells into the uterus, selection of residency sites, and proliferation in situ are independent of NOS-2. By contrast, their differentiation, including transcription and translation of the cytotoxic molecule perforin, was shown to rely on normal expression of the NOS-2 gene.

Interferon-γ–induced membrane PAF-receptor expression confers tumor cell susceptibility to NK perforin-dependent lysis

Perforin is known to display a membranolytic activity on tumor cells. Nevertheless, perforin release during natural killer (NK)–cell activation is not sufficient to induce membrane target-cell damage. On the basis of the ability of perforin to interact with phospholipids containing a choline phosphate headgroup, we identify the platelet-activating factor (PAF) and its membrane receptor as crucial components in tumor cell killing activity of human resting NK cells. We demonstrate for the first time that upon activation, naive NK cells release the choline phosphate–containing lysolipid PAF, which binds to perforin and acts as an agonist on perforin-induced membrane damage. PAF is known to incorporate cell membranes using a specific receptor. Here we show that interferon-γ (IFN–γ) secreted from activated NK cells ends in PAF-receptor expression on perforin-sensitive K562 cells but not on perforin-resistant Daudi cells. In order to prove the capacity of PAF to interact simultaneously with its membrane PAF receptor and with perforin, we successfully co-purified the 3 components in the presence of bridging PAF molecules. The functional activity of this complex was further examined. The aim was to determine whether membrane PAF-receptor expression on tumor cells, driven to express this receptor, could render them sensitive to the perforin lytic pathway. The results confirmed that transfection of the PAF-receptor complementary DNA into major histocompatibility complex class I and Fas-receptor negative tumor cells restored susceptibility to naive NK cells and perforin attack. Failure of IFN-γ to induce membrane PAF receptor constitutes the first described mechanism for tumor cells to resist the perforin lytic pathway.

Interferon-γ–induced membrane PAF-receptor expression confers tumor cell susceptibility to NK perforin-dependent lysis

Perforin is known to display a membranolytic activity on tumor cells. Nevertheless, perforin release during natural killer (NK)–cell activation is not sufficient to induce membrane target-cell damage. On the basis of the ability of perforin to interact with phospholipids containing a choline phosphate headgroup, we identify the platelet-activating factor (PAF) and its membrane receptor as crucial components in tumor cell killing activity of human resting NK cells. We demonstrate for the first time that upon activation, naive NK cells release the choline phosphate–containing lysolipid PAF, which binds to perforin and acts as an agonist on perforin-induced membrane damage. PAF is known to incorporate cell membranes using a specific receptor. Here we show that interferon-γ (IFN–γ) secreted from activated NK cells ends in PAF-receptor expression on perforin-sensitive K562 cells but not on perforin-resistant Daudi cells. In order to prove the capacity of PAF to interact simultaneously with its membrane PAF receptor and with perforin, we successfully co-purified the 3 components in the presence of bridging PAF molecules. The functional activity of this complex was further examined. The aim was to determine whether membrane PAF-receptor expression on tumor cells, driven to express this receptor, could render them sensitive to the perforin lytic pathway. The results confirmed that transfection of the PAF-receptor complementary DNA into major histocompatibility complex class I and Fas-receptor negative tumor cells restored susceptibility to naive NK cells and perforin attack. Failure of IFN-γ to induce membrane PAF receptor constitutes the first described mechanism for tumor cells to resist the perforin lytic pathway.

Granzyme A initiates an alternative pathway for granule-mediated apoptosis

Granzyme (gzm) B-deficient cytotoxic lymphocytes (CTL) have a severe defect in the rapid induction of target cell apoptosis that is almost completely corrected by prolonged incubation of the CTL effectors and their targets. We show in this report that perforin-dependent, gzmB-independent cytotoxicity is caused by gzmA (or tightly linked genes). CTL deficient for gzmA and gzmB retain normal perforin function, but these CTL have a cytotoxic defect in vivo that is as severe as perforin-deficient CTL. Collectively, these results suggest that perforin provides target cell access and/or trafficking signals for the gzms, and that the gzms themselves deliver the lethal hits. The gzmA pathway appears to function independently from gzmB and may therefore provide a critical "back-up" system when gzmB is inhibited in the target cell.

IL-18 Up-Regulates Perforin-Mediated NK Activity Without Increasing Perforin Messenger RNA Expression by Binding to Constitutively Expressed IL-18 Receptor

IL-18 is a powerful inducer of IFN-γ production, particularly in collaboration with IL-12. IL-18, like IL-12, also augments NK activity. Here we investigated the molecular mechanism underlying the up-regulation of killing activity of NK cells by IL-18. IL-18, like IL-12, dose dependently enhanced NK activity of splenocytes. This action was further enhanced by costimulation with IL-12. Treatment with anti-IL-2R Ab did not affect IL-18- and/or IL-12-augmented NK activity, and splenocytes from IFN-γ-deficient mice showed enhanced NK activity following stimulation with IL-12 and/or IL-18. Splenocytes from the mice deficient in both IL-12 and IL-18 normally responded to IL-18 and/or IL-12 with facilitated NK activity, suggesting that functional NK cells develop in the absence of IL-12 and IL-18. IL-18R, as well as IL-12R mRNA, was constitutively expressed in splenocytes from SCID mice, which lack T cells and B cells but have intact NK cells, and in those from IL-12 and IL-18 double knockout mice. NK cells isolated from SCID splenocytes expressed IL-18R on their surface. IL-18, in contrast to IL-12, did not enhance mRNA expression of perforin, a key molecule for exocytosis-mediated cytotoxicity. However, pretreatment with concanamycin A completely inhibited this IL-18- and/or IL-12-augmented NK activity. Furthermore, IL-18, like IL-12, failed to enhance NK activity of splenocytes from perforin-deficient mice. These data suggested that NK cells develop and express IL-12R and IL-18R in the absence of IL-12 or IL-18, and that both IL-18 and IL-12 directly and independently augment perforin-mediated cytotoxic activity of NK cells.

Degranulation plays an essential part in regulating cell surface expression of Fas ligand in T cells and natural killer cells

Fas ligand (FasL) triggers apoptosis during cytotoxicity mediated by cytotoxic T lymphocytes and during immune downregulation. The ability of T cells and natural killer cells to trigger apoptosis through this mechanism is controlled by the cell surface expression of FasL (ref. 2). Because FasL expression is up-regulated on activation, FasL was thought to be delivered directly to the cell surface. Here we show that newly synthesized FasL is stored in specialized secretory lysosomes in both CD4+ and CD8+ T cells and natural killer cells, and that polarized degranulation controls the delivery of FasL to the cell surface. In this way, FasL-mediated apoptosis is finely controlled by receptor-mediated target-cell recognition. The cytoplasmic tail of FasL contains signals that sort FasL to secretory lysosomes in hemopoietic cells. This pathway may provide a general mechanism for controlling the cell surface appearance of proteins involved in immune regulation.

Lipid binding ridge on loops 2 and 3 of the C2A domain of synaptotagmin I as revealed by NMR spectroscopy

The C2A domain of synaptotagmin I, which binds Ca2+ and anionic phospholipids, serves as a Ca2+ sensor during excitation-secretion coupling. We have used multidimensional NMR to locate the region of C2A from rat synaptotagmin I that interacts, in the presence of Ca2+, with phosphatidylserine. Untagged, recombinant C2A was double-labeled with 13C and 15N, and triple-resonance NMR data were collected from C2A samples containing either Ca2+ alone or Ca2+ plus 6:0 phosphatidylserine. Phospholipid binding led to changes in chemical shifts of backbone atoms in residues Arg233 and Phe234 of loop 3 (a loop that also binds Ca2+) and His198, Val205, and Phe206 of loop 2. These residues lie along a straight line on a surface ridge of the C2A domain. The only other residue that exhibited appreciable chemical shift changes upon adding lipid was His254; however, because His254 is located on the other side of the molecule from the phospholipid docking site defined by the other residues, its shifts may result from nonspecific interactions. The results show that the "docking ridge" responsible for Ca2+-dependent membrane association is localized on the opposite side of the C2A domain from the transmembrane and C2B domains of synaptotagmin.

Mechanism of phospholipid binding by the C2A-domain of synaptotagmin I

Synaptotagmin I is a synaptic vesicle membrane protein that probably functions as a Ca2+ sensor in neurotransmitter release and contains two C2-domains which bind Ca2+. The first C2-domain of synaptotagmin I (the C2A-domain) binds phospholipids in a Ca2+-dependent manner similar to that of the C2-domains of protein kinase C, cytoplasmic phospholipase A2, and phospholipase Cdelta1. Although the tertiary structure of these C2-domains is known, the molecular basis for their Ca2+-dependent interactions with phospholipids is unclear. We have now investigated the mechanisms involved in Ca2+-dependent phospholipid binding by the C2A-domain of synaptotagmin I. Our data show that the C2A-domain binds negatively charged liposomes in an electrostatic interaction that is determined by the charge density of the liposome surface but not by the phospholipid headgroup. At the tip of the C2A-domain, three tightly clustered Ca2+-binding sites are formed by five aspartates and one serine. Mutations in these aspartate and serine residues demonstrated that all three Ca2+-binding sites are required for phospholipid binding. The Ca2+ binding sites at the top of the C2A-domain are surrounded by positively charged amino acids that were shown by mutagenesis to be also involved in phospholipid binding. Our results yield a molecular picture of the interactions between a C2-domain and phospholipids. Binding is highly electrostatic and occurs between the surfaces of the phospholipid bilayer and of the tip of the C2A-domain. The data suggest that the negatively charged phospholipid headgroups interact with the basic side chains surrounding the Ca2+-binding sites and with bound Ca2+ ions, thereby filling empty coordination sites and increasing the apparent affinity for Ca2+. In addition, insertion of hydrophobic side chains may contribute to phospholipid binding. This model is likely to be general for other C2-domains, with the relative contributions of electrostatic and hydrophobic interactions dictated by the exposed side chains surrounding the Ca2+-binding region.

Interaction between a Ca2+-binding protein calreticulin and perforin, a component of the cytotoxic T-cell granules

Calreticulin is a component of cytotoxic T-lymphocyte and NK lymphocyte granules. We report here that granule-associated calreticulin terminates with the KDEL endoplasmic reticulum retrieval amino acid sequence and somehow escapes the KDEL retrieval system. In perforin knock-out mice calreticulin is still targeted into the granules. Thus, calreticulin will traffic without perforin to cytotoxic granules. In the granules, calreticulin and perforin are associated as documented by (i) copurification of calreticulin with perforin but not with granzymes and (ii) immunoprecipitation of a calreticulin-perforin complex using specific antibodies. By using calreticulin affinity chromatography and protein ligand blotting we show that perforin binds to calreticulin in the absence of Ca2+ and the two proteins dissociate upon exposure to 0.1 mM or higher Ca2+ concentration. Perforin interacts strongly with the P-domain of calreticulin (the domain which has high Ca2+-binding affinity and chaperone function) as revealed by direct protein-protein interaction, ligand blotting, and the yeast two-hybrid techniques. Our results suggest that calreticulin may act as Ca2+-regulated chaperone for perforin. This action will serve to protect the CTL during biogenesis of granules and may also serve to regulate perforin lytic action after release.

Ca2+ binding to synaptotagmin: how many Ca2+ ions bind to the tip of a C2-domain?

C2-domains are widespread protein modules with diverse Ca2+-regulatory functions. Although multiple Ca2+ ions are known to bind at the tip of several C2-domains, the exact number of Ca2+-binding sites and their functional relevance are unknown. The first C2-domain of synaptotagmin I is believed to play a key role in neurotransmitter release via its Ca2+-dependent interactions with syntaxin and phospholipids. We have studied the Ca2+-binding mode of this C2-domain as a prototypical C2-domain using NMR spectroscopy and site-directed mutagenesis. The C2-domain is an elliptical module composed of a beta-sandwich with a long axis of 50 A. Our results reveal that the C2-domain binds three Ca2+ ions in a tight cluster spanning only 6 A at the tip of the module. The Ca2+-binding region is formed by two loops whose conformation is stabilized by Ca2+ binding. Binding involves one serine and five aspartate residues that are conserved in numerous C2-domains. All three Ca2+ ions are required for the interactions of the C2-domain with syntaxin and phospholipids. These results support an electrostatic switch model for C2-domain function whereby the beta-sheets of the domain provide a fixed scaffold for the Ca2+-binding loops, and whereby interactions with target molecules are triggered by a Ca2+-induced switch in electrostatic potential.

Direct interaction of a Ca2+-binding loop of synaptotagmin with lipid bilayers

Synaptotagmin 1 binds Ca2+ and membranes via its C2A-domain and plays an essential role in excitation-secretion coupling. In this study, we sought to identify Ca2+- and membrane-induced local conformational changes in the C2A-domain of synaptotagmin and to delineate the C2A-lipid binding interface. To address these questions native phenylalanine residues were replaced, at each face of the domain, with tryptophan reporters. Changes in tryptophanyl fluorescence indicated that Ca2+ induced long range conformational changes throughout C2A, including regions distant from an established Ca2+-binding site. Addition of liposomes resulted in Ca2+-dependent increases in the fluorescence of tryptophans 193, 231, and 234. Only the tryptophan residues at positions 234 and 231, which lie within a Ca2+-binding loop of C2A, exhibited liposome-induced blue shifts in their emission spectra. Quenching experiments, using membrane-imbedded doxyl spin labels, revealed that tryptophan residues 231 and 234 penetrated lipid bilayers. These data delineate the lipid binding interface of C2A and provide the first evidence for adjacent Ca2+- and lipid-binding sites within a C2-domain. The penetration of C2A into membranes may function to bring components of the fusion machinery into contact with the lipid bilayer to initiate exocytosis.